Skip to content

Commit

Permalink
Only save the dataset_mix_dir in the main process
Browse files Browse the repository at this point in the history
  • Loading branch information
vwxyzjn committed Sep 19, 2024
1 parent 97c9963 commit 1fdc9ca
Show file tree
Hide file tree
Showing 2 changed files with 4 additions and 4 deletions.
4 changes: 2 additions & 2 deletions open_instruct/dpo_tune.py
Original file line number Diff line number Diff line change
Expand Up @@ -577,7 +577,7 @@ def main(args: FlatArguments):
args.dataset_mixer,
configs=args.dataset_config_name,
splits=["train"],
save_data_dir=args.dataset_mix_dir,
save_data_dir=args.dataset_mix_dir if accelerator.is_main_process else None,
columns_to_keep=["chosen", "rejected"],
)
elif args.dataset_mixer_list is not None:
Expand All @@ -586,7 +586,7 @@ def main(args: FlatArguments):
args.dataset_mixer_list,
configs=args.dataset_config_name,
splits=["train"],
save_data_dir=args.dataset_mix_dir,
save_data_dir=args.dataset_mix_dir if accelerator.is_main_process else None,
columns_to_keep=["chosen", "rejected"],
)
else:
Expand Down
4 changes: 2 additions & 2 deletions open_instruct/finetune.py
Original file line number Diff line number Diff line change
Expand Up @@ -526,7 +526,7 @@ def main(args: FlatArguments):
args.dataset_mixer,
configs=args.dataset_config_name,
splits=["train"],
save_data_dir=args.dataset_mix_dir,
save_data_dir=args.dataset_mix_dir if accelerator.is_main_process else None,
columns_to_keep=["messages"],
)
elif args.dataset_mixer_list is not None:
Expand All @@ -535,7 +535,7 @@ def main(args: FlatArguments):
args.dataset_mixer_list,
configs=args.dataset_config_name,
splits=["train"],
save_data_dir=args.dataset_mix_dir,
save_data_dir=args.dataset_mix_dir if accelerator.is_main_process else None,
columns_to_keep=["messages"],
)
else:
Expand Down

0 comments on commit 1fdc9ca

Please sign in to comment.